DOCSLIB.ORG

Stable Distributions Models for Heavy Tailed Data

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Stable Distributions Models for Heavy Tailed Data Stable Distributions Models for Heavy Tailed Data John P. Nolan [email protected] Math/Stat Department American University Copyright ⃝c 2017 John P. Nolan Processed January 30, 2018 Dedicated to Martha, Julia and Erin and Anne Zemitus Nolan (1919-2016) Contents I Univariate Stable Distributions 1 1 Basic Properties of Univariate Stable Distributions 3 1.1 Definition of stable . 4 1.2 Other definitions of stability . 7 1.3 Parameterizations of stable laws . 7 1.4 Densities and distribution functions . 12 1.5 Tail probabilities, moments and quantiles . 14 1.6 Sums of stable random variables . 18 1.7 Simulation . 21 1.8 Generalized Central Limit Theorem . 22 1.9 Problems . 23 2 Modeling with Stable Distributions 25 2.1 Lighthouse problem . 26 2.2 Distribution of masses in space . 27 2.3 Random walks . 28 2.4 Hitting time for Brownian motion . 33 2.5 Differential equations and fractional diffusions . 33 2.6 Economic applications . 35 2.6.1 Stock returns . 35 2.6.2 Foreign exchange rates . 35 2.6.3 Value-at-risk . 35 2.6.4 Other economic applications . 36 2.6.5 Long tails in business, political science, and medicine . 36 iv Contents 2.6.6 Multiple assets . 37 2.7 Time series . 38 2.8 Signal processing . 38 2.9 Embedding of Banach spaces . 39 2.10 Stochastic resonance . 39 2.11 Miscellaneous applications . 40 2.11.1 Gumbel copula . 40 2.11.2 Exponential power distributions . 40 2.11.3 Queueing theory . 40 2.11.4 Geology . 41 2.11.5 Physics . 42 2.11.6 Hazard function, survival analysis and reliability . 42 2.11.7 Network traffic . 44 2.11.8 Computer Science . 44 2.11.9 Biology and medicine . 45 2.11.10 Discrepancies . 45 2.11.11 Punctuated change . 45 2.11.12 Central Pre-Limit Theorem . 45 2.11.13 Extreme values models . 46 2.12 Behavior of the sample mean and variance . 46 2.13 Appropriateness of infinite variance models . 48 2.14 Historical notes . 51 2.15 Problems . 51 3 Technical Results for Univariate Stable Distributions 53 3.1 Proofs of Basic Theorems of Chapter 1 . 53 3.1.1 Stable distributions as infinitely divisible distributions . 63 3.2 Densities and distribution functions . 64 3.2.1 Series expansions . 74 3.2.2 Modes . 75 3.2.3 Duality . 80 3.3 Numerical algorithms . 82 3.3.1 Computation of distribution functions and densities . 82 3.3.2 Spline approximation of densities . 83 3.3.3 Simulation . 84 3.4 More on parameterizations . 86 3.5 Tail behavior . 90 3.6 Moments and other transforms . 103 3.7 Convergence of stable laws in terms of (a;b;g;d) . 111 3.8 Combinations of stable random variables . 113 3.9 Distributions derived from stable distributions . 121 3.9.1 Log-stable . 121 3.9.2 Exponential stable . 121 3.9.3 Amplitude of a stable random variable . 122 3.9.4 Ratios of stable terms . 123 3.9.5 Wrapped stable distribution . 125 Contents v 3.9.6 Discretized stable distributions . 125 3.10 Stable distributions arising as functions of other distributions . 126 3.11 Extreme value distributions and Tweedie distributions . 127 3.11.1 Stable mixtures of extreme value distributions . 127 3.11.2 Tweedie distributions . 129 3.12 Stochastic series representations . 129 3.13 Generalized Central Limit Theorem and Domains of Attraction . 130 3.14 Central Pre-Limit Theorem . 138 3.15 Entropy . 138 3.16 Differential equations and stable semi-groups . 139 3.17 Problems . 142 4 Univariate Estimation 147 4.1 Order statistics . 147 4.2 Tail based estimation . 148 4.2.1 Hill estimator . 150 4.3 Extreme value theory estimate of a . 152 4.4 Quantile based estimation . 153 4.5 Characteristic function based estimation . 157 4.6 Moment based methods of estimation . 159 4.7 Maximum likelihood estimation . 160 4.7.1 Asymptotic normality and Fisher information matrix . 162 4.7.2 The score function . 165 4.8 Other methods of estimation . 168 4.8.1 Log absolute value estimation . 168 4.8.2 U statistic based estimation . 168 4.8.3 Minimum distance estimator . 169 4.8.4 Conditional maximum likelihood estimation . 170 4.8.5 Miscellaneous methods . 170 4.9 Comparisons of estimators . 171 4.10 Assessing a stable fit . 172 4.10.1 Likelihood ratio tests and goodness-of-fit tests . 173 4.10.2 Testing the stability hypothesis . 174 4.10.3 Diagnostics . 174 4.11 Applications . 176 4.12 Fitting stable distributions to concentration data . 184 4.13 Estimation for discretized stable distributions . 184 4.14 Discussion . 185 4.15 Problems . 185 II Multivariate Stable Distributions 187 5 Basic Properties of Multivariate Stable Distributions 189 5.1 Definition of jointly stable . 189 5.2 Representations of jointly stable vectors . 192 vi Contents 5.2.1 Projection based representation . 193 5.2.2 Spectral measure representation . 195 5.2.3 Stable stochastic integral representation . 198 5.2.4 Stochastic series representation . 199 5.2.5 Zonoid representation . 200 5.3 Multivariate stable densities and probabilities . 202 5.3.1 Multivariate tail probabilities . 204 5.4 Examples of multivariate stable distributions . 205 5.4.1 Independent components . 205 5.4.2 Discrete spectral measures . 205 5.4.3 Radially symmetric and elliptically contoured stable laws . 209 5.4.4 Symmetric stable laws . 212 5.4.5 Sub-stable laws and linear combinations . 213 5.4.6 Complexity of the joint dependence structure . 213 5.5 Sums of stable random vectors . 216 5.6 Simulation . 218 5.7 Multivariate generalized central limit theorem . 219 5.8 Problems . 219 6 Technical Results for Multivariate Stable Distributions 223 6.1 Proofs of basic properties of multivariate stable distributions . 223 6.2 Parameterizations . ..
Load more

Recommended publications
  • Factor of Safety and Probability of Failure
    Factor of safety and probability of failure Introduction How does one assess the acceptability of an engineering design? Relying on judgement alone can lead to one of the two extremes illustrated in Figure 1. The first case is economically unacceptable while the example illustrated in the drawing on the right violates all normal safety standards. Figure 1: Rockbolting alternatives involving individual judgement. (Drawings based on a cartoon in a brochure on rockfalls published by the Department of Mines of Western Australia.) Sensitivity studies The classical approach used in designing engineering structures is to consider the relationship between the capacity C (strength or resisting force) of the element and the demand D (stress or disturbing force). The Factor of Safety of the structure is defined as F = C/D and failure is assumed to occur when F is less than unity. Factor of safety and probability of failure Rather than base an engineering design decision on a single calculated factor of safety, an approach which is frequently used to give a more rational assessment of the risks associated with a particular design is to carry out a sensitivity study. This involves a series of calculations in which each significant parameter is varied systematically over its maximum credible range in order to determine its influence upon the factor of safety. This approach was used in the analysis of the Sau Mau Ping slope in Hong Kong, described in detail in another chapter of these notes. It provided a useful means of exploring a range of possibilities and reaching practical decisions on some difficult problems.
    [Show full text]
  • Probabilistic Stability Analysis
    TABLE OF CONTENTS Page A-7 Probabilistic Limit State Analysis ....................................................... A-7-1 A-7.1 Key Concepts ....................................................................... A-7-1 A-7.2 Example: FOSM and MC Analysis for Heave at the Toe of a Levee ................................................................. A-7-4 A-7.3 Example: RCC Gravity Dam Stability .............................. A-7-14 A-7.4 Example: Screening-Level Check of Embankment Post-Liquefaction Stability ............................................ A-7-19 A-7.5 Example: Foundation Rock Wedge Stability .................... A-7-23 A-7.6 Model Uncertainty ............................................................. A-7-26 A-7.7 References .......................................................................... A-7-28 Tables Page A-7-1 Variables in Levee Heave Analysis ................................................. A-7-7 A-7-2 FOSM Calculations for Water Surface at Levee Crest .................... A-7-8 A-7-3 Variable Distributions for MC Simulation .................................... A-7-11 A-7-4 Comparison of MC and FOSM Analysis Results .......................... A-7-12 A-7-5 Correlation Coefficients for Water Surface at the Levee Crest ..... A-7-13 A-7-6 Comparison of MC and FOSM Sensitivity Analysis Results ........ A-7-13 A-7-7 Summary of Concrete Input Properties .......................................... A-7-16 A-7-8 RCC Dam Sensitivity Rankings..................................................... A-7-18 A-7-9 Summary
    [Show full text]
  • Estimation of Α-Stable Sub-Gaussian Distributions for Asset Returns
    Estimation of α-Stable Sub-Gaussian Distributions for Asset Returns Sebastian Kring, Svetlozar T. Rachev, Markus Hochst¨ otter¨ , Frank J. Fabozzi Sebastian Kring Institute of Econometrics, Statistics and Mathematical Finance School of Economics and Business Engineering University of Karlsruhe Postfach 6980, 76128, Karlsruhe, Germany E-mail: [email protected] Svetlozar T. Rachev Chair-Professor, Chair of Econometrics, Statistics and Mathematical Finance School of Economics and Business Engineering University of Karlsruhe Postfach 6980, 76128 Karlsruhe, Germany and Department of Statistics and Applied Probability University of California, Santa Barbara CA 93106-3110, USA E-mail: [email protected] Markus Hochst¨ otter¨ Institute of Econometrics, Statistics and Mathematical Finance School of Economics and Business Engineering University of Karlsruhe Postfach 6980, 76128, Karlsruhe, Germany Frank J. Fabozzi Professor in the Practice of Finance School of Management Yale University New Haven, CT USA 1 Abstract Fitting multivariate α-stable distributions to data is still not feasible in higher dimensions since the (non-parametric) spectral measure of the characteristic func- tion is extremely difficult to estimate in dimensions higher than 2. This was shown by Chen and Rachev (1995) and Nolan, Panorska and McCulloch (1996). α-stable sub-Gaussian distributions are a particular (parametric) subclass of the multivariate α-stable distributions. We present and extend a method based on Nolan (2005) to estimate the dispersion matrix of an α-stable sub-Gaussian dis- tribution and estimate the tail index α of the distribution. In particular, we de- velop an estimator for the off-diagonal entries of the dispersion matrix that has statistical properties superior to the normal off-diagonal estimator based on the covariation.
    [Show full text]
  • The Holtsmark-Continuum Model for the Statistical Description of a Plasma
    On fluctuations in plasmas : the Holtsmark-continuum model for the statistical description of a plasma Citation for published version (APA): Dalenoort, G. J. (1970). On fluctuations in plasmas : the Holtsmark-continuum model for the statistical description of a plasma. Technische Hogeschool Eindhoven. https://doi.org/10.6100/IR108607 DOI: 10.6100/IR108607 Document status and date: Published: 01/01/1970 Document Version: Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication: • A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website. • The final author version and the galley proof are versions of the publication after peer review. • The final published version features the final layout of the paper including the volume, issue and page numbers. Link to publication General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.
    [Show full text]
  • 1 Stable Distributions in Streaming Computations
    1 Stable Distributions in Streaming Computations Graham Cormode1 and Piotr Indyk2 1 AT&T Labs – Research, 180 Park Avenue, Florham Park NJ, [email protected] 2 Laboratory for Computer Science, Massachusetts Institute of Technology, Cambridge MA, [email protected] 1.1 Introduction In many streaming scenarios, we need to measure and quantify the data that is seen. For example, we may want to measure the number of distinct IP ad- dresses seen over the course of a day, compute the difference between incoming and outgoing transactions in a database system or measure the overall activity in a sensor network. More generally, we may want to cluster readings taken over periods of time or in different places to find patterns, or find the most similar signal from those previously observed to a new observation. For these measurements and comparisons to be meaningful, they must be well-defined. Here, we will use the well-known and widely used Lp norms. These encom- pass the familiar Euclidean (root of sum of squares) and Manhattan (sum of absolute values) norms. In the examples mentioned above—IP traffic, database relations and so on—the data can be modeled as a vector. For example, a vector representing IP traffic grouped by destination address can be thought of as a vector of length 232, where the ith entry in the vector corresponds to the amount of traffic to address i. For traffic between (source, destination) pairs, then a vec- tor of length 264 is defined. The number of distinct addresses seen in a stream corresponds to the number of non-zero entries in a vector of counts; the differ- ence in traffic between two time-periods, grouped by address, corresponds to an appropriate computation on the vector formed by subtracting two vectors, and so on.
    [Show full text]
  • Hyperbolicity and Stable Polynomials in Combinatorics and Probability
    Hyperbolicity and stable polynomials in combinatorics and probability Robin Pemantle 1,2 ABSTRACT: These lectures survey the theory of hyperbolic and stable polynomials, from their origins in the theory of linear PDE’s to their present uses in combinatorics and probability theory. Keywords: amoeba, cone, dual cone, G˚arding-hyperbolicity, generating function, half-plane prop- erty, homogeneous polynomial, Laguerre–P´olya class, multiplier sequence, multi-affine, multivariate stability, negative dependence, negative association, Newton’s inequalities, Rayleigh property, real roots, semi-continuity, stochastic covering, stochastic domination, total positivity. Subject classification: Primary: 26C10, 62H20; secondary: 30C15, 05A15. 1Supported in part by National Science Foundation grant # DMS 0905937 2University of Pennsylvania, Department of Mathematics, 209 S. 33rd Street, Philadelphia, PA 19104 USA, pe- [email protected] Contents 1 Introduction 1 2 Origins, definitions and properties 4 2.1 Relation to the propagation of wave-like equations . 4 2.2 Homogeneous hyperbolic polynomials . 7 2.3 Cones of hyperbolicity for homogeneous polynomials . 10 3 Semi-continuity and Morse deformations 14 3.1 Localization . 14 3.2 Amoeba boundaries . 15 3.3 Morse deformations . 17 3.4 Asymptotics of Taylor coefficients . 19 4 Stability theory in one variable 23 4.1 Stability over general regions . 23 4.2 Real roots and Newton’s inequalities . 26 4.3 The Laguerre–P´olya class . 31 5 Multivariate stability 34 5.1 Equivalences . 36 5.2 Operations preserving stability . 38 5.3 More closure properties . 41 6 Negative dependence 41 6.1 A brief history of negative dependence . 41 6.2 Search for a theory . 43 i 6.3 The grail is found: application of stability theory to joint laws of binary random variables .
    [Show full text]
  • Geometric Stable Laws Through Series Representations
    Serdica Math. J. 25 (1999), 241-256 GEOMETRIC STABLE LAWS THROUGH SERIES REPRESENTATIONS Tomasz J. Kozubowski, Krzysztof Podg´orski Communicated by S. T. Rachev Abstract. Let (Xi) be a sequence of i.i.d. random variables, and let N be a geometric random variable independent of (Xi). Geometric stable distributions are weak limits of (normalized) geometric compounds, SN = X + + XN , when the mean of N converges to infinity. By an appro- 1 · · · priate representation of the individual summands in SN we obtain series representation of the limiting geometric stable distribution. In addition, we [Nt] study the asymptotic behavior of the partial sum process SN (t) = Xi, i=1 and derive series representations of the limiting geometric stable processP and the corresponding stochastic integral. We also obtain strong invariance principles for stable and geometric stable laws. 1. Introduction. An increasing interest has been seen recently in geo- metric stable (GS) distributions: the class of limiting laws of appropriately nor- malized random sums of i.i.d. random variables, (1) S = X + + X , N 1 · · · N 1991 Mathematics Subject Classification: 60E07, 60F05, 60F15, 60F17, 60G50, 60H05 Key words: geometric compound, invariance principle, Linnik distribution, Mittag-Leffler distribution, random sum, stable distribution, stochastic integral 242 Tomasz J. Kozubowski, Krzysztof Podg´orski where the number of terms is geometrically distributed with mean 1/p, and p 0 → (see, e.g., [7], [8], [10], [11], [12], [13], [14] and [21]). These heavy-tailed distribu- tions provide useful models in mathematical finance (see, e.g., [1], [20], [13]), as well as in a variety of other fields (see, e.g., [6] for examples, applications, and extensive references for geometric compounds (1)).
    [Show full text]
  • Extreme Value Theory (Or How to Go Beyond of the Range Data)
    Extreme Value Theory (or how to go beyond of the range data) Sept 2007, Romania Philippe Naveau Laboratoire des Sciences du Climat et l’Environnement (LSCE) Gif-sur-Yvette, France Katz et al., Statistics of extremes in hydrology, Advances in Water Resources 25 (2002) 1287-1304 Motivation Univariate EVT Non-stationary extremes Spatial extremes Conclusions Extreme quotes 1 “Man can believe the impossible, but man can never believe the improbable” Oscar Wilde (Intentions, 1891) 2 “Il est impossible que l’improbable n’arrive jamais” Emil Julius Gumbel (1891-1966) Extreme events ? ... a probabilistic concept linked to the tail behavior : low frequency of occurrence, large uncertainty and sometimes strong amplitude. Region of interest Motivation Univariate EVT Non-stationary extremes Spatial extremes Conclusions Important issues in Extreme Value Theory Applied statistics An asymptotic probabilistic concept Non-stationarity A statistical modeling approach Identifying clearly assumptions Univariate Multivariate Non-parametric Parametric Assessing uncertainties Goodness of fit and model selection Independence Theoritical probability Motivation Univariate EVT Non-stationary extremes Spatial extremes Conclusions Outline 1 Motivation Heavy rainfalls Three applications 2 Univariate EVT Asymptotic result Historical perspective GPD Parameters estimation Brief summary of univariate iid EVT 3 Non-stationary extremes Spatial interpolation of return levels Downscaling of heavy rainfalls 4 Spatial extremes Assessing spatial dependences among maxima 5 Conclusions
    [Show full text]
  • Mixed-Stable Models: an Application to High-Frequency Financial Data
    entropy Article Mixed-Stable Models: An Application to High-Frequency Financial Data Igoris Belovas 1,* , Leonidas Sakalauskas 2, Vadimas Starikoviˇcius 3 and Edward W. Sun 4 1 Institute of Data Science and Digital Technologies, Faculty of Mathematics and Informatics, Vilnius University, LT-04812 Vilnius, Lithuania 2 Department of Information Technologies, Faculty of Fundamental Sciences, Vilnius Gediminas Technical University, LT-2040 Vilnius, Lithuania; [email protected] 3 Department of Mathematical Modelling, Faculty of Fundamental Sciences, Vilnius Gediminas Technical University, LT-2040 Vilnius, Lithuania; [email protected] 4 KEDGE Business School, Accounting, Finance, & Economics Department (CFE), Campus Bordeaux, 33405 Talence, France; [email protected] * Correspondence: [email protected] Abstract: The paper extends the study of applying the mixed-stable models to the analysis of large sets of high-frequency financial data. The empirical data under review are the German DAX stock index yearly log-returns series. Mixed-stable models for 29 DAX companies are constructed employing efficient parallel algorithms for the processing of long-term data series. The adequacy of the modeling is verified with the empirical characteristic function goodness-of-fit test. We propose the smart-D method for the calculation of the a-stable probability density function. We study the impact of the accuracy of the computation of the probability density function and the accuracy of ML-optimization on the results of the modeling and processing time. The obtained mixed-stable parameter estimates can be used for the construction of the optimal asset portfolio. Citation: Belovas, I.; Sakalauskas, L.; Starikoviˇcius,V.; Sun, E.W. Keywords: mixed-stable models; high-frequency data; stock index returns Mixed-Stable Models: An Application to High-Frequency Financial Data.
    [Show full text]
  • 9 Sums of Independent Random Variables
    STA 711: Probability & Measure Theory Robert L. Wolpert 9 Sums of Independent Random Variables We continue our study of sums of independent random variables, Sn = X1 + + Xn. If E 2 E 2··· each Xi is square-integrable, with mean µi = Xi and variance σi = [(Xi µi) ], then Sn is E V − 2 2 square integrable too with mean Sn = µ n := i n µi and variance Sn = σ n := i n σi . ≤ ≤ ≤ ≤ But what about the actual probability distribution? If the Xi have density functions fi(xi) P P then Sn has a density function too; for example, with n = 2, S2 = X1 + X2 has CDF F (s) and pdf f(s)= F ′(s) given by P[S2 s]= F (s) = f1(x1)f2(x2) dx1dx2 ≤ x1+x2 s ZZ ≤ s x2 ∞ − = f1(x1)f2(x2) dx1dx2 Z−∞ Z−∞ ∞ ∞ = F (s x )f (x ) dx = F (s x )f (x ) dx 1 − 2 2 2 2 2 − 1 1 1 1 Z−∞ Z−∞ ∞ ∞ f(s)= F ′(s) = f (s x )f (x ) dx = f (x )f (s x ) dx , 1 − 2 2 2 2 1 1 2 − 1 1 Z−∞ Z−∞ the convolution f = f1 ⋆ f2 of f1(x1) and f2(x2). Even if the distributions aren’t abso- lutely continuous, so no pdfs exist, S2 has a distribution measure µ given by µ(ds) = R µ1(dx1)µ2(ds x1). There is an analogous formula for n = 3, but it is quite messy; things get worse and worse− as n increases, so this is not a promising approach for studying the R distribution of sums Sn for large n.
    [Show full text]
  • A Bayesian Hierarchical Generalized Pareto Spatial Model for Exceedances Over a Threshold
    A Bayesian Hierarchical Generalized Pareto Spatial Model for Exceedances Over a Threshold Fernando Ferraz do Nascimento Universidade Federal do Piau´ı, Campus Petronio Portella, CCN II, 64000 Teresina-PI, Brazil. Bruno Sanso´ University of California, Santa Cruz, 1156 High Street, SOE 2, Santa Cruz, CA-95064, USA. Summary. Extreme value theory focuses on the study of rare events and uses asymptotic properties to estimate their associated probabilities. Easy availability of georeferenced data has prompted a growing interest in the analysis of spatial extremes. Most of the work so far has focused on models that can handle block maxima, with few examples of spatial models for exceedances over a threshold. Using a hierarchical representation, we propose a spatial process, that has generalized Pareto distributions as its marginals. The process is in the domain of attraction of the max-stable family. It has the ability to capture both, asymptotic dependence and independence. We use a Bayesian approach for inference of the process parameters that can be efficiently applied to a large number of spatial locations. We assess the flexibility of the model and the accuracy of the inference by considering some simulated examples. We illustrate the model with an analysis of data for temperature and rainfall in California. Keywords: Generalized Pareto distribution; Max-Stable process; Bayesian hierarchical model; Spatial extremes; MCMC; Asymptotic Dependence 1. Introduction The statistical analysis of extreme values focuses on inference for rare events that cor- respond to the tails of probability distributions. As such, it is a key ingredient in the assessment of the risk of phenomena that can have strong societal impacts like floods, heat waves, high concentration of pollutants, crashes in the financial markets, among oth- ers.
    [Show full text]
  • Fractional Poisson Process in Terms of Alpha-Stable Densities
    FRACTIONAL POISSON PROCESS IN TERMS OF ALPHA-STABLE DENSITIES by DEXTER ODCHIGUE CAHOY Submitted in partial fulfillment of the requirements For the degree of Doctor of Philosophy Dissertation Advisor: Dr. Wojbor A. Woyczynski Department of Statistics CASE WESTERN RESERVE UNIVERSITY August 2007 CASE WESTERN RESERVE UNIVERSITY SCHOOL OF GRADUATE STUDIES We hereby approve the dissertation of DEXTER ODCHIGUE CAHOY candidate for the Doctor of Philosophy degree * Committee Chair: Dr. Wojbor A. Woyczynski Dissertation Advisor Professor Department of Statistics Committee: Dr. Joe M. Sedransk Professor Department of Statistics Committee: Dr. David Gurarie Professor Department of Mathematics Committee: Dr. Matthew J. Sobel Professor Department of Operations August 2007 *We also certify that written approval has been obtained for any proprietary material contained therein. Table of Contents Table of Contents . iii List of Tables . v List of Figures . vi Acknowledgment . viii Abstract . ix 1 Motivation and Introduction 1 1.1 Motivation . 1 1.2 Poisson Distribution . 2 1.3 Poisson Process . 4 1.4 α-stable Distribution . 8 1.4.1 Parameter Estimation . 10 1.5 Outline of The Remaining Chapters . 14 2 Generalizations of the Standard Poisson Process 16 2.1 Standard Poisson Process . 16 2.2 Standard Fractional Generalization I . 19 2.3 Standard Fractional Generalization II . 26 2.4 Non-Standard Fractional Generalization . 27 2.5 Fractional Compound Poisson Process . 28 2.6 Alternative Fractional Generalization . 29 3 Fractional Poisson Process 32 3.1 Some Known Properties of fPp . 32 3.2 Asymptotic Behavior of the Waiting Time Density . 35 3.3 Simulation of Waiting Time . 38 3.4 The Limiting Scaled nth Arrival Time Distribution .
    [Show full text]